Artificial Intelligence in Ecuadorian Higher Education: A PRISMA-Based Systematic Literature Review of Benefits, Challenges, and Research Gaps

Artificial intelligence (AI) is transforming higher education; however, evidence of its research and application in Ecuador remains fragmented. We systematically reviewed the internationally indexed literature on AI in Ecuadorian higher education. Following the PRISMA 2020 guidelines, Scopus and the Web of Science Core Collection were searched for peer-reviewed publications from 2015 to 2025. Of the 271 unique records screened, 62 studies met the eligibility criteria; they were appraised using design-specific instruments (MMAT 2018, PROBAST) and analyzed through thematic synthesis across eight axes. The corpus concentrates on 2024–2025 (71%) and is dominated by learning analytics (16 studies) and generative AI (14 studies), followed by perception studies (10 studies). Reported benefits—early warning, efficiency, assessment support, and skill development—rest on technical performance, usability, and satisfaction rather than on measured learning or institutional gains. Persistent challenges include infrastructure constraints, uneven digital competencies, ethical and privacy concerns, and methodologically fragile local evidence; 30 of the 59 appraised studies were rated as having high concern. Ethics, governance, and academic integrity were rarely the primary research objects (two of 62 studies), although 28 raised them as secondary findings. This review derives a research agenda prioritizing governance, pedagogical theorization, methodological rigor, and outcome evidence.